Computational Cost Efficient Model of Losses for Multi-port Active-bridge Converters

Multi-port active-bridge (MAB) converters, in single and clustered architecture, are used for many different applications. Due to intertwining of currents and powers of all their ports, finding a closed form solution for optimal control of single and clustered MAB converters that leads to highest efficiency is not an option. Therefore, optimization of power distribution in a cluster of MAB converters must be done by trial and error, which requires a model that can predict the losses. A detailed and accurate model of losses, however, would need large amounts of memory and processing resources, which makes real-time optimization of control and modulations costly, if not almost impossible. This work seeks to propose a computational cost efficient model, which is able to predict the losses with acceptable accuracy. Given the operating point of a MAB converter, this model is able to predict total losses using polynomial calculations. This efficient model can be employed for optimizing control of a cluster of MAB converters. Predictions of this model are compared against experimental measurements to validate its correct operation.